CN103763346B - A distributed resource scheduling method and device - Google Patents

Abstract

A kind of distributed resource scheduling method and device.This method includes：The first standard deviation of the resource utilization of the same type business virtual machine in distributed cluster system in each server is obtained according to predetermined obtaining step, when first standard deviation is more than or equal to scheduling thresholds, the same type business virtual machine balance of any server in distributed cluster system is dispatched in other servers of distributed cluster system, again the second standard deviation of the resource utilization of the same type business virtual machine in each server after balance is dispatched is obtained according to predetermined obtaining step, the second standard deviation obtained more at least one times after balance scheduling, perform the corresponding balance scheduling of the second minimum standard deviation.By according in distributed cluster system in each server business virtual machine type, or the resource ratio of server where being accounted for according to all kinds of business virtual machines in each server carries out distributed resource scheduling, it is ensured that the service quality of each business virtual machine in each server.

Description

A distributed resource scheduling method and device

technical field

The present invention relates to resource scheduling technology, in particular to a distributed resource scheduling method and device.

Background technique

In virtualization or cloud computing solutions, each server in a server cluster generally includes multiple virtual machines running different types of services. When different types of services are deployed on the same server, that is, a computing node, for example, a computing node simultaneously If there are desktop cloud virtual machines and compilation business virtual machines, it is necessary to consider how to ensure the respective service quality, ensure that the resource occupation of each virtual machine does not interfere with each other, and maximize the use of cluster resources.

In the existing technology, Quality of Service (QoS) technology is used to ensure the service quality of different types of services on the same computing node, which requires the total amount and proportion of different types of virtual machines to remain stable. For example, a computing node has 40 Virtual machines, including 15 desktop cloud virtual machines and 25 compiled cloud virtual machines, ensure that high-share virtual machines can obtain a considerable share of resources by setting the share of virtual machines to high, medium and low; using Distributed Resource Scheduler (Distributed Resource Scheduler, DRS), to maximize resource utilization in the cluster. For example, DRS migrates the overloaded virtual machine to a server with low load. At this time, the resource occupancy rate of the overloaded virtual machine is improved, and the original server other The resource occupancy rate of the virtual machine has also been improved, and the overall resource occupancy rate of the cluster has been improved.

However, after starting DRS, if the virtual machine is migrated to another server, the proportion of different types of business virtual machines in the server will be unbalanced. Therefore, after starting DRS, how to ensure the service quality of each business virtual machine in each server is currently a problem. problems that urgently need to be resolved.

Contents of the invention

Embodiments of the present invention provide a distributed resource scheduling method and device, which can ensure the service quality of each service virtual machine in each server during distributed resource scheduling.

In the first aspect, a distributed resource scheduling method is provided, including:

Obtain the resource utilization rate of the same type of business virtual machines in each server in the distributed cluster system;

According to the resource utilization rate of the same type of service virtual machine in each server, the first standard deviation of the resource utilization rate of the same type of service virtual machine in each server is obtained through a predetermined acquisition step, and the acquisition step is: Obtain the average resource usage rate of the same type of service virtual machines in each server according to the resource usage rate of the same type of service virtual machines in each server, and obtain the resource usage rate of the same type of service virtual machines in each server rate and the average resource usage rate, and obtain the first standard deviation of the resource usage rates of the same type of service virtual machines in each server;

When the first standard deviation of the resource usage rates of the same type of service virtual machines in each server is greater than or equal to the scheduling threshold of the resource usage rate of the same type of service virtual machines in each server, the distributed cluster The same type of business virtual machine of any server in the system is balancedly dispatched to other servers in the distributed cluster system;

Obtain the resource utilization rate of the same type of business virtual machines in each server in the distributed cluster system after balanced scheduling;

According to the resource utilization rate of the same type of service virtual machine in each server in the distributed cluster system after the balanced scheduling, obtain the resource utilization rate of the same type of service virtual machine in each server through the predetermined acquisition step the second standard deviation of

Comparing the second standard deviation obtained after the balanced scheduling at least once, and executing the balanced scheduling corresponding to the smallest second standard deviation.

In a first possible implementation manner, the method further includes comparing the second standard deviation obtained after the balanced scheduling at least once, and before executing the balanced scheduling corresponding to the smallest second standard deviation. include:

If the second standard deviation of the resource utilization rate of the same type of service virtual machine in each server after the next balanced scheduling is greater than or equal to the resource utilization rate of the same type of service virtual machine in each server after the current balanced scheduling The second standard deviation of the second standard deviation, or as of the next balance scheduling, if the cumulative number of balanced scheduling exceeds the first set threshold, the acquisition step and the next balanced scheduling will be stopped.

In the second aspect, a distributed resource scheduling method is provided, including:

According to the proportion of various service virtual machines in each server in the distributed cluster system to the resources of each server and the resource utilization rate of various service virtual machines in each server, the resource utilization rate of each server is obtained;

According to the resource utilization rate of each server, the third standard deviation of the resource utilization rate of the distributed cluster system is acquired through a predetermined acquisition step, and the acquisition step is: acquiring the resource utilization rate of each server according to the the average resource usage rate of the distributed cluster system, and according to the resource usage rate of each server and the average resource usage rate of the distributed cluster system, obtain the third standard deviation of the resource usage rate of the distributed cluster system;

When the third standard deviation of the resource utilization rate of the distributed cluster system is greater than or equal to the scheduling threshold of the resource utilization rate of the distributed cluster system, any service virtual machine in each server is balancedly scheduled to the In other servers of the distributed cluster system;

Obtain the resource utilization rate of each server in the distributed cluster system after balanced scheduling;

According to the resource usage rate of each server in the distributed cluster system after balanced scheduling, the fourth standard deviation of the resource usage rate of the distributed cluster system after balanced scheduling is obtained through the predetermined acquisition step;

Comparing the fourth standard deviation acquired after the balanced scheduling at least once, and executing the balanced scheduling corresponding to the smallest fourth standard deviation.

In a first possible implementation manner, before comparing the fourth standard deviation acquired after the balanced scheduling at least once, before executing the balanced scheduling corresponding to the smallest fourth standard deviation, the method further include:

If the fourth standard deviation of the resource usage rate of the distributed cluster system after the next balanced scheduling is greater than or equal to the fourth standard deviation of the resource usage rate of the distributed cluster system after the current balanced scheduling, or until the next Balanced scheduling, the cumulative number of times the balanced scheduling is performed exceeds the first set threshold, or after the service virtual machine is balancedly scheduled from any first server to any second server, any one of the first servers or all The change of the total amount of various types of business virtual machines in any second server is greater than the second set threshold, and/or the proportion of various types of business virtual machines in any one of the first servers or the second server is If the resource change of any one of the first servers is greater than the third set threshold, the execution of the acquisition step and the next balance scheduling will be stopped.

In a third aspect, a distributed resource scheduling device is provided, including:

The first obtaining unit is used to obtain the resource utilization rate of the same type of business virtual machines in each server in the distributed cluster system;

The second acquisition unit is configured to acquire the first standard deviation of the resource usage rates of the same type of service virtual machines in each server through a predetermined acquisition method according to the resource usage rates of the same type of service virtual machines in each server. , the obtaining method is: obtaining the average resource utilization rate of the same type of service virtual machine in each server according to the resource utilization rate of the same type of service virtual machine in each server, and according to the same resource utilization rate in each server The resource utilization rate of the type of service virtual machine and the average resource utilization rate are obtained by obtaining the first standard deviation of the resource utilization rate of the same type of service virtual machine in each server;

The first scheduling unit is configured to: when the first standard deviation of the resource usage rates of the same type of service virtual machines in each server is greater than or equal to the scheduling threshold of the resource usage rate of the same type of service virtual machines in each server , balancing and scheduling the same type of service virtual machine of any server in the distributed cluster system to other servers in the distributed cluster system;

The third obtaining unit is used to obtain the resource utilization rate of the same type of service virtual machine in each server in the distributed cluster system after balanced scheduling;

The fourth acquisition unit is configured to acquire the same type of service virtual machines in each server through the predetermined acquisition method according to the resource utilization rate of the same type of service virtual machines in each server in the distributed cluster system after the balanced scheduling. The second standard deviation of the resource utilization rate of the business virtual machine;

a first comparison unit, configured to compare the second standard deviation obtained after at least one balance scheduling;

A first executing unit, configured to execute the balanced scheduling corresponding to the smallest second standard deviation.

In a first possible implementation manner, the first execution unit is further configured to:

If the second standard deviation of the resource utilization rate of the same type of service virtual machine in each server after the next balanced scheduling is greater than or equal to the resource utilization rate of the same type of service virtual machine in each server after the current balanced scheduling The second standard deviation of the second standard deviation, or as of the next balance scheduling, if the cumulative number of balanced scheduling exceeds the first set threshold, stop executing the second standard of the resource utilization rate of the same type of service virtual machine in each server Poor fetch and next balanced dispatch.

In a fourth aspect, a distributed resource scheduling device is provided, including:

The fifth acquiring unit is configured to acquire the resource utilization rate of each type of business virtual machine in each server according to the proportion of each type of business virtual machine in each server in the distributed cluster system to the resources of each server. server resource usage;

The sixth acquisition unit is configured to acquire the third standard deviation of the resource utilization rate of the distributed cluster system through a predetermined acquisition method according to the resource utilization rate of each server, and the acquisition method is: according to the resource utilization rate of each server Obtain the average resource utilization rate of the distributed cluster system, and obtain the resource utilization rate of the distributed cluster system according to the resource utilization rate of each server and the average resource utilization rate of the distributed cluster system The third standard deviation of the rate;

The second scheduling unit is configured to schedule any service in each server when the third standard deviation of the resource utilization rate of the distributed cluster system is greater than or equal to the scheduling threshold The virtual machine is balanced and dispatched to other servers in the distributed cluster system;

The seventh obtaining unit is used to obtain the resource utilization rate of each server in the distributed cluster system after balanced scheduling;

An eighth acquisition unit, configured to acquire the fourth criterion of the resource utilization rate of the distributed cluster system after balanced scheduling through the predetermined acquisition method according to the resource utilization rate of each server in the distributed cluster system after balanced scheduling Difference;

a second comparison unit, configured to compare the fourth standard deviation obtained after at least one balance scheduling;

The second executing unit is configured to execute the balanced scheduling corresponding to the smallest fourth standard deviation.

In a first possible implementation manner, the second execution unit is further configured to:

If the fourth standard deviation of the resource usage rate of the distributed cluster system after the next balanced scheduling is greater than or equal to the fourth standard deviation of the resource usage rate of the distributed cluster system after the current balanced scheduling, or until the next Balanced scheduling, the cumulative number of times the balanced scheduling is performed exceeds the first set threshold, or after the service virtual machine is balancedly scheduled from any first server to any second server, any one of the first servers or all The change of the total amount of various types of business virtual machines in any second server is greater than the second set threshold, and/or the proportion of various types of business virtual machines in any one of the first servers or the second server is If the resource change of any one of the first servers is greater than the third set threshold, the acquisition of the resource utilization rate of the distributed cluster system after the balanced scheduling and the next balanced scheduling are stopped.

By adopting the technical solution of a distributed resource scheduling method and device provided by the embodiment of the present invention, according to the types of business virtual machines in each server in the distributed cluster system, or according to the types of business virtual machines in each server occupying the server where they are located, Distributed resource scheduling based on the resource ratio can guarantee the service quality of each business virtual machine in each server.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. Those skilled in the art can also obtain other drawings based on these drawings without creative work.

FIG. 1 is a schematic flowchart of a distributed resource scheduling method provided by an embodiment of the present invention;

FIG. 2 is a schematic flowchart of another distributed resource scheduling method provided by an embodiment of the present invention;

FIG. 3 is a schematic flowchart of another distributed resource scheduling method provided by an embodiment of the present invention;

FIG. 4 is a schematic flowchart of another distributed resource scheduling method provided by an embodiment of the present invention;

FIG. 5 is a schematic structural diagram of a distributed resource scheduling device provided by an embodiment of the present invention;

FIG. 6 is a schematic structural diagram of another distributed resource scheduling device provided by an embodiment of the present invention;

FIG. 7 is a schematic structural diagram of a distributed resource scheduling device provided by an embodiment of the present invention;

FIG. 8 is a schematic structural diagram of another distributed resource scheduling device provided by an embodiment of the present invention.

detailed description

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

Please refer to FIG. 1 , which is a schematic flowchart of a distributed resource scheduling method provided by an embodiment of the present invention. The method includes the following steps:

Step S101 , acquiring resource usage rates of service virtual machines of the same type in each server in the distributed cluster system.

A virtualized or cloud computing server cluster includes multiple servers, and each server includes multiple virtual machines, and these multiple virtual machines can run different types of services, which are called service virtual machines. DRS technology is used to allocate resources in server clusters. The system allocates certain resources to each server, and each server allocates corresponding resources to its virtual machines. Therefore, the server cluster is also called a distributed cluster system. The resources here Refers to control, storage, and input/output resources such as the central processing unit (Central Processing Unit, CPU), memory, and input/output (Input/Output, I/O).

In this step, the resource utilization rate of the same type of business virtual machine in each server in the distributed cluster system can be obtained. The resource usage rate of the machine can be reported to DRS, or the DRS can request the resource usage rate of this type of virtual machine from the monitor, and the resource usage rate of this type of business virtual machine can be obtained from the monitor for one time or multiple times . Here, the resource usage rate may be the usage rate of a certain resource, or may be a weighted value of resource usage rates of various resources such as CPU, memory, and I/O.

Each server in the distributed cluster system has one or more types of business virtual machines, for example, there are desktop cloud virtual machines and compilation cloud virtual machines. This embodiment obtains the resources of the same type of virtual machines in each server according to the business type. usage rate.

For example, there are three servers in the cluster. Server 1 has 2 desktop cloud virtual machines and 2 compiled cloud virtual machines. Server 2 has 3 desktop cloud virtual machines and 2 compiled cloud virtual machines. Server 3 has desktop cloud. 1 virtual machine and 2 compiled cloud virtual machines. DRS allocates certain CPU, memory, and I/O resources to servers 1, 2, and 3 in the cluster, and each server allocates certain resources to various types of virtual machines inside it. CPU, memory, and I/O resources. Taking CPU resource allocation as an example, DRS allocates CPU resources of quantity A to server 1, allocates CPU resources of quantity B to server 2, and allocates CPU resources of quantity C to server 3. , Server 1 allocates CPU resources of A1, A2, A3, and A4 to the 2 desktop cloud virtual machines and 2 compiled cloud virtual machines respectively. Taking the resource usage rate of the desktop cloud virtual machine as an example, the 2 desktop cloud virtual machines in server 1 The resource utilization rate of the desktop cloud virtual machine is 30% and 40% respectively, then the resource utilization rate of the CPU of the desktop cloud virtual machine in server 1 is (30%*A1+40%*A2)/(A1+A2), Then calculate the total memory and I/O resource usage of the desktop cloud virtual machine in Server 1 in the same way, and then perform weighted calculations on the total CPU, memory, and I/O resource usage to obtain Server 1 The resource usage of the desktop cloud virtual machine, and so on to obtain the resource usage of the desktop cloud virtual machines of servers 2 and 3. At the same time, the resource utilization rate of the compiled cloud virtual machine in each server can also be obtained in the same manner.

Step S102, according to the resource utilization rate of the same type of service virtual machine in each server, obtain the first standard deviation of the resource utilization rate of the same type of service virtual machine in each server through a predetermined acquisition step, the acquisition The steps are: according to the resource usage of the same type of service virtual machines in each server, the average resource usage of the same type of service virtual machines in each server is obtained, and according to the resource usage of the same type of service virtual machines in each of the servers and the average resource usage rate, and obtain the first standard deviation of the resource usage rates of the same type of service virtual machines in each server.

The DRS technology is adopted to ensure the maximum utilization of resources in the cluster. Therefore, the standard deviation between the resource utilization of the same type of service virtual machines in each server in the cluster and the average resource utilization of the same type of service virtual machines in each server is calculated. The standard deviation is used to reflect the degree of dispersion of the data, and its calculation formula is shown in formula (1):

In formula (1), x _i is the resource utilization rate of the same type of business virtual machine in each server; μ is the average resource utilization rate of the same type of service virtual machine in each server, N is the number of servers in the cluster, σ is the standard deviation.

According to the step of obtaining the standard deviation, it is first necessary to obtain the average resource utilization rate of the same type of service virtual machines in each server. In step S101, the resource utilization rate of the same type of service virtual machine in each server has been obtained, and the average resource utilization rate of the same type of service virtual machine in each server can be obtained according to the existing calculation method. For example, in the preceding example, if there are three servers, the obtained resource utilization rates of the same type of service virtual machines in each server are summed and then divided by the number of servers to obtain the average resource utilization rate.

Then, according to the formula (1), the standard deviation of the resource utilization rate of the same type of service virtual machines in each server is obtained.

Step S103, when the first standard deviation of the resource utilization rate of the same type of service virtual machine in each server is greater than or equal to the scheduling threshold of the resource utilization rate of the same type of service virtual machine in each server, set the The same type of service virtual machine of any server in the distributed cluster system is balancedly scheduled to other servers in the distributed cluster system.

By comparing the obtained standard deviation with the scheduling threshold of the resource usage rate of the same type of business virtual machines in each server, the system can judge whether the resource usage of this type of business virtual machines in each server is unbalanced. The larger the standard deviation, the higher the imbalance rate. higher. It should be noted that since there are multiple types of business virtual machines in each server, the standard deviation of the obtained average resource usage rate is calculated for each same type of business virtual machine in each server, and the obtained standard deviation The result may be that the resource usage of a certain type of service virtual machine is balanced, while the resource usage of another type of service virtual machine is unbalanced.

When the first standard deviation of resource usage rates of service virtual machines of the same type in each server is greater than or equal to a scheduling threshold of resource usage rates of service virtual machines of the same type in each server, scheduling is triggered. The scheduling process may follow recursive scheduling. In the recursive scheduling process, the same type of service virtual machine of any server in the distributed cluster system is balancedly scheduled to other servers in the distributed cluster system.

Step S104, obtaining the resource utilization rate of the same type of service virtual machines in each server in the distributed cluster system after balanced scheduling.

Step S105, according to the resource utilization rate of the same type of service virtual machines in each server in the distributed cluster system after the balanced scheduling, obtain the information of the same type of service virtual machines in each server through the predetermined acquisition step The second standard deviation of resource usage.

After each round of balanced scheduling, the influence of the scheduled virtual machine on the imbalance rate is obtained, that is, the standard deviation of the resource utilization rate of the same type of service virtual machines in each server after the balanced scheduling is obtained again.

Since the resource utilization rate of the same type of service virtual machine in each server changes after each round of balanced scheduling, it is first necessary to obtain the resource utilization rate of the same type of service virtual machine in each server in the distributed cluster system after balanced scheduling , and then obtain the standard deviation of the resource utilization rate of the same type of service virtual machines in each server after the round of balanced scheduling according to the predetermined standard deviation acquisition step.

Step S106, comparing the second standard deviation obtained after the balanced scheduling at least once, and executing the balanced scheduling corresponding to the smallest second standard deviation.

Perform multiple rounds of the above-mentioned balanced scheduling. During the balanced scheduling, certain strategies can be adopted for scheduling, such as scheduling heavy-duty business virtual machines to other servers. This embodiment of the present invention does not limit it. During multiple rounds of the above-mentioned balanced scheduling, record Each round of balanced scheduling obtains the standard deviation of the resource utilization rate of the same type of business virtual machine in each server after the round of balanced scheduling, and then compares multiple standard deviations obtained after multiple rounds of balanced scheduling to select the smallest standard deviation. Finally, the balanced scheduling corresponding to the minimum standard deviation is executed, that is, which virtual machine(s) are scheduled to which server(s) the balanced scheduling based on the minimum standard deviation is performed, and then the balanced scheduling is performed.

It should be noted that since the distributed resource scheduling is performed for the same type of service virtual machines in each server, the distributed resource scheduling of each type of service virtual machine in each server can be performed separately. If the resources of a certain type of service virtual machine When balancing is used, the distributed resource scheduling is performed.

It can be seen that according to a distributed resource scheduling method provided by an embodiment of the present invention, by performing distributed resource scheduling according to the types of service virtual machines in each server in a distributed cluster system, the service quality of each service virtual machine in each server can be guaranteed. .

Please refer to FIG. 2 , which is a schematic flowchart of another distributed resource scheduling method provided by an embodiment of the present invention. The difference between this method and the previous embodiment is that it also includes step S206: if the second standard deviation of the resource utilization rate of the same type of service virtual machine in each server is greater than or equal to the current balance after the next balance scheduling After scheduling, the second standard deviation of the resource utilization rate of the same type of service virtual machine in each server, or as of the next balanced scheduling, if the cumulative number of times the balanced scheduling is performed exceeds the first set threshold, then stop executing the Get step and next balance schedule.

Using recursive scheduling can make the scheduling efficiency the highest, and set the recursive exit condition in the recursive scheduling, that is, if the second standard deviation of the resource utilization rate of the same type of service virtual machine in each server is greater than or equal to the next balanced scheduling The second standard deviation of the resource utilization rate of the same type of service virtual machine in each server after the current balanced scheduling is performed, or until the next balanced scheduling, the cumulative number of times the balanced scheduling is performed exceeds the first set threshold, then stop Execute the acquisition step and the next balance scheduling. This helps to stop the above acquisition steps and balance scheduling in time, obtain the minimum standard deviation, and execute the balance scheduling corresponding to the minimum standard deviation.

Please refer to FIG. 3 , which is a schematic flowchart of another distributed resource scheduling method provided by an embodiment of the present invention. The method includes the following steps:

Step S301, according to the ratio of various types of business virtual machines in each server in the distributed cluster system to the resources of each server and the resource utilization rate of each type of business virtual machines in each server, obtain the resource usage of each server Rate.

Each server in the distributed cluster system has one or more types of business virtual machines, for example, there are desktop cloud virtual machines and compilation cloud virtual machines, and DRS allocates certain computing, storage, and input/output to each server in the cluster resources, and each server allocates the resources it obtains to various types of business virtual machines inside it according to a certain proportion. Ensuring that the total amount and ratio of resources of different types of business virtual machines in each server remain stable can ensure the service quality of various types of business virtual machines in each server.

For example, there are three servers in the cluster. Server 1 has 2 desktop cloud virtual machines and 2 compiled cloud virtual machines. Server 2 has 3 desktop cloud virtual machines and 2 compiled cloud virtual machines. Server 3 has desktop cloud. 1 virtual machine, 2 compiled cloud virtual machines, DRS allocates certain CPU, memory and I/O resources to servers 1, 2, and 3 in the cluster, and each server allocates certain CPU to each internal virtual machine , memory and I/O resources, taking CPU resource allocation as an example, DRS allocates CPU resources of quantity A to server 1, allocates CPU resources of quantity B to server 2, and allocates CPU resources of quantity C to server 3, and server 1 Allocate the CPU resources of A1, A2, A3 and A4 to the 2 desktop cloud virtual machines and 2 compiled cloud virtual machines respectively, and the resource utilization rates of the 2 desktop cloud virtual machines in server 1 are 30% and 40% respectively. %, the resource utilization rates of the two compiled cloud virtual machines in server 1 are 40% and 50% respectively, then the CPU resource utilization rate of server 1 is obtained (30%*A1+40%*A2+40%*A3+ 50%A4)/(A1+A2+A3+A4), and then calculate the memory and I/O resource usage of various virtual machines in server 1 in the same way, and then calculate the CPU, memory and I/O resource usage of server 1 The resource usage of I/O is weighted and calculated to obtain the resource usage of server 1, and so on for servers 2 and 3.

Step S302, according to the resource utilization rate of each server, obtain the third standard deviation of the resource utilization rate of the distributed cluster system through a predetermined acquisition step, the acquisition step is: according to the resource utilization rate of each server Obtaining the average resource usage rate of the distributed cluster system, and according to the resource usage rate of each server and the average resource usage rate of the distributed cluster system, obtaining a third value of the resource usage rate of the distributed cluster system standard deviation.

The difference from the foregoing embodiments is that since it is aimed at various types of business virtual machines in each server, the resource utilization rate of each server and the average resource utilization rate of the entire distributed cluster system are obtained in the same manner as the foregoing embodiments, and then , obtain the standard deviation of the resource utilization rate of the distributed cluster system according to the formula (1).

Step S303, when the third standard deviation of the resource usage rate of the distributed cluster system is greater than or equal to the scheduling threshold of the resource usage rate of the distributed cluster system, balance scheduling of any service virtual machine in each server to other servers in the distributed cluster system.

Compare the obtained standard deviation of the resource utilization rate of the distributed cluster system with the scheduling threshold of the resource utilization rate of the distributed cluster system, if the standard deviation of the resource utilization rate of the distributed cluster system is greater than or equal to the distributed cluster system The scheduling threshold of the resource usage of the system triggers balanced scheduling. The object of balanced scheduling can be any service virtual machine in each server.

Step S304, obtaining the resource utilization rate of each server in the distributed cluster system after balanced scheduling.

Step S305, according to the resource usage rate of each server in the distributed cluster system after balanced scheduling, obtain the fourth standard deviation of the resource usage rate of the distributed cluster system after balanced scheduling through the predetermined acquisition step.

Step S306, comparing the fourth standard deviation obtained after the balanced scheduling at least once, and executing the balanced scheduling corresponding to the smallest fourth standard deviation.

Step S304-Step S306 are similar to Step S104-Step S106 of the previous embodiment, and will not be repeated here. The difference is that, for various types of business virtual machines in each server, the obtained is the distributed cluster system Standard deviation of resource usage.

It can be seen that, according to a distributed resource scheduling method provided by an embodiment of the present invention, by performing distributed resource scheduling according to the resource ratio of various business virtual machines in each server to the server where it is located, the service of each business virtual machine in each server can be guaranteed. quality.

Please refer to FIG. 4 , which is a schematic flowchart of another distributed resource scheduling method provided by an embodiment of the present invention. The method is different from the foregoing embodiments in that it further includes step S406: if the fourth standard deviation of the resource utilization rate of the distributed cluster system after the next balanced scheduling is greater than or equal to the distribution after the current balanced scheduling The fourth standard deviation of the resource utilization rate of the cluster system, or until the next balance scheduling, the cumulative number of balanced scheduling exceeds the first set threshold, or the service virtual machine is balanced from any first server After going to any second server, the change of the total amount of all kinds of business virtual machines in the any first server or the any second server is greater than the second set threshold, and/or the any first server If the change in the resources of any first server occupied by various types of service virtual machines in the server or any second server is greater than a third set threshold, the execution of the acquisition step and the next balance scheduling will be stopped.

Using recursive scheduling can make the scheduling efficiency the highest, and set the recursive exit condition in the recursive scheduling, that is, if the fourth standard deviation of the resource utilization rate of the distributed cluster system after the next balanced scheduling is greater than or equal to the current balanced scheduling The fourth standard deviation of the resource utilization rate of the distributed cluster system, or as of the next balanced scheduling, the cumulative number of times the balanced scheduling is performed exceeds the first set threshold, or the service virtual machine is switched from any first After the server is balanced and dispatched to any second server, the change of the total amount of various business virtual machines in the any first server or the any second server is greater than the second set threshold, and/or the any If the resource change of any one of the first servers occupied by various types of business virtual machines in the first server or the second server is greater than the third set threshold, stop performing the acquisition step and the next balancing scheduling. This helps to stop the above acquisition steps and balance scheduling in time, obtain the minimum standard deviation, and execute the balance scheduling corresponding to the minimum standard deviation.

Please refer to FIG. 5 , which is a schematic structural diagram of a distributed resource scheduling device provided by an embodiment of the present invention. The device 1000 includes:

The first acquiring unit 11 is configured to acquire the resource utilization rate of the same type of service virtual machines in each server in the distributed cluster system.

A virtualized or cloud computing server cluster includes multiple servers, and each server includes multiple virtual machines, and these multiple virtual machines can run different types of services, which are called service virtual machines. DRS technology is used to allocate resources in server clusters. The system allocates certain resources to each server, and each server allocates corresponding resources to its virtual machines. Therefore, the server cluster is also called a distributed cluster system. The resources here Refers to control, storage, and input/output resources such as the central processing unit (Central Processing Unit, CPU), memory, and input/output (Input/Output, I/O).

The first obtaining unit 11 obtains the resource usage rate of the same type of service virtual machine in each server in the distributed cluster system. The resource usage rate of the machine can be reported to DRS, or the DRS can request the resource usage rate of this type of virtual machine from the monitor, and the resource usage rate of this type of business virtual machine can be obtained from the monitor for one time or multiple times . Here, the resource usage rate may be the usage rate of a certain resource, or may be a weighted value of resource usage rates of various resources such as CPU, memory, and I/O.

Each server in the distributed cluster system has one or more types of business virtual machines, for example, there are desktop cloud virtual machines and compilation cloud virtual machines. This embodiment obtains the resources of the same type of virtual machines in each server according to the business type. usage rate.

For example, there are three servers in the cluster. Server 1 has 2 desktop cloud virtual machines and 2 compiled cloud virtual machines. Server 2 has 3 desktop cloud virtual machines and 2 compiled cloud virtual machines. Server 3 has desktop cloud. 1 virtual machine and 2 compiled cloud virtual machines. DRS allocates certain CPU, memory, and I/O resources to servers 1, 2, and 3 in the cluster, and each server allocates certain resources to various types of virtual machines inside it. CPU, memory, and I/O resources. Taking CPU resource allocation as an example, DRS allocates CPU resources of quantity A to server 1, allocates CPU resources of quantity B to server 2, and allocates CPU resources of quantity C to server 3. , Server 1 allocates CPU resources of A1, A2, A3, and A4 to the 2 desktop cloud virtual machines and 2 compiled cloud virtual machines respectively. Taking the resource usage rate of the desktop cloud virtual machine as an example, the 2 desktop cloud virtual machines in server 1 The resource utilization rate of the desktop cloud virtual machine is 30% and 40% respectively, then the resource utilization rate of the CPU of the desktop cloud virtual machine in server 1 is (30%*A1+40%*A2)/(A1+A2), Then calculate the total memory and I/O resource usage of the desktop cloud virtual machine in Server 1 in the same way, and then perform weighted calculations on the total CPU, memory, and I/O resource usage to obtain Server 1 The resource usage of the desktop cloud virtual machine, and so on to obtain the resource usage of the desktop cloud virtual machines of servers 2 and 3. At the same time, the resource utilization rate of the compiled cloud virtual machine in each server can also be obtained in the same manner.

The second acquisition unit 12 is configured to acquire the first standard of the resource usage rate of the same type of service virtual machine in each server through a predetermined acquisition method according to the resource usage rate of the same type of service virtual machine in each server Poor, the acquisition method is as follows: Obtain the average resource utilization rate of the same type of service virtual machine in each server according to the resource utilization rate of the same type of service virtual machine in each server, and according to the resource utilization rate of the same type of service virtual machine in each server The resource utilization rate of the same type of service virtual machine and the average resource utilization rate obtain the first standard deviation of the resource utilization rate of the same type of service virtual machine in each server.

The DRS technology is adopted to ensure the maximum utilization rate of resources in the cluster. Therefore, the second acquisition unit 12 acquires the resource utilization rate of the same type of service virtual machines in each server in the cluster and the average resource utilization of the same type of service virtual machines in each server. The standard deviation of the rate, the standard deviation is used to reflect the degree of dispersion of the data, and its calculation formula is shown in formula (1).

According to the formula for calculating the standard deviation, the second obtaining unit 12 first needs to obtain the average resource utilization rate of the same type of service virtual machines in each server. The first obtaining unit 11 has obtained the resource utilization rate of the same type of service virtual machine in each server, and the second obtaining unit 12 can obtain the average resource utilization rate of the same type of service virtual machine in each server according to the existing calculation method. For example, in the preceding example, if there are three servers, the obtained resource utilization rates of the same type of service virtual machines in each server are summed and then divided by the number of servers to obtain the average resource utilization rate.

Then, the second acquiring unit 12 acquires the standard deviation of the resource utilization rates of the same type of service virtual machines in each server according to the formula (1).

The first scheduling unit 13 is configured to: when the first standard deviation of the resource usage rates of the same type of service virtual machines in each server is greater than or equal to the scheduling threshold of the resource usage rate of the same type of service virtual machines in each server , the same type of service virtual machine of any server in the distributed cluster system is balancedly scheduled to other servers in the distributed cluster system.

By comparing the obtained standard deviation with the scheduling threshold of the resource usage rate of the same type of business virtual machines in each server, the system can judge whether the resource usage of this type of business virtual machines in each server is unbalanced. The larger the standard deviation, the higher the imbalance rate. higher. It should be noted that since there are multiple types of business virtual machines in each server, the standard deviation of the obtained average resource usage rate is calculated for each same type of business virtual machine in each server, and the obtained standard deviation The result may be that the resource usage of a certain type of service virtual machine is balanced, while the resource usage of another type of service virtual machine is unbalanced.

When the first standard deviation of the resource usage rates of the same type of service virtual machines in each server is greater than or equal to the scheduling threshold of the resource usage rates of the same type of service virtual machines in each server, the first scheduling unit 13 triggers scheduling. The scheduling process may follow recursive scheduling. In the recursive scheduling process, the same type of service virtual machine of any server in the distributed cluster system is balancedly scheduled to other servers in the distributed cluster system.

The third obtaining unit 14 is configured to obtain the resource utilization rate of the same type of service virtual machines in each server in the distributed cluster system after balanced scheduling.

The fourth acquisition unit 15 is configured to acquire the same type of service virtual machine in each server in the predetermined acquisition method according to the resource utilization rate of the same type of service virtual machine in each server in the distributed cluster system after the balanced scheduling. The second standard deviation of the resource usage of type business virtual machines.

After each round of balanced scheduling, the fourth acquisition unit 15 acquires the influence of the scheduled virtual machine on the unbalance rate, that is, the fourth acquisition unit 15 reacquires the standard deviation of the resource utilization rate of the same type of service virtual machines in each server after the balanced scheduling.

Since the resource utilization rate of the same type of service virtual machines in each server changes after each round of balanced scheduling, first the third acquisition unit 14 needs to obtain the same type of service virtual machines in each server in the distributed cluster system after balanced scheduling. Then the fourth acquisition unit 15 acquires the standard deviation of the resource utilization rate of the same type of service virtual machines in each server after the round of balance scheduling according to the predetermined standard deviation acquisition method.

A first comparison unit 16, configured to compare the second standard deviation obtained after at least one balance scheduling;

The first executing unit 17 is configured to execute the balanced scheduling corresponding to the smallest second standard deviation.

Perform multiple rounds of the above-mentioned balanced scheduling. During the balanced scheduling, certain strategies can be adopted for scheduling, such as scheduling heavy-duty business virtual machines to other servers. This embodiment of the present invention does not limit it. During multiple rounds of the above-mentioned balanced scheduling, record Each round of balanced scheduling obtains the standard deviation of the resource utilization rate of the same type of business virtual machine in each server after the round of balanced scheduling, and then compares multiple standard deviations obtained after multiple rounds of balanced scheduling to select the smallest standard deviation. The first execution unit 17 finally executes the balanced scheduling corresponding to the minimum standard deviation, that is, which or which virtual machines are scheduled to which or which servers are scheduled for the balanced scheduling performed by obtaining the minimum standard deviation, then the balanced scheduling is executed .

It should be noted that since the distributed resource scheduling is performed for the same type of service virtual machines in each server, the distributed resource scheduling of each type of service virtual machine in each server can be performed separately. If the resources of a certain type of service virtual machine When balancing is used, the distributed resource scheduling is performed.

It can be seen that, according to a distributed resource scheduling device provided by an embodiment of the present invention, by performing distributed resource scheduling according to the types of service virtual machines in each server in a distributed cluster system, the service quality of each service virtual machine in each server can be guaranteed. .

As an implementation manner, the first execution unit 17 may also be configured to if the second standard deviation of the resource utilization rates of the same type of service virtual machines in each server after the next balanced scheduling is greater than or equal to the current balanced scheduling If the second standard deviation of the resource utilization rate of the same type of service virtual machine in each server, or as of the next balance scheduling, the cumulative number of times the balance scheduling is performed exceeds the first set threshold, then the execution of each server will be stopped. Acquisition of the second standard deviation of the resource utilization rate of the same type of business virtual machines within the same type of service and the next balance scheduling.

Using recursive scheduling can make the scheduling efficiency the highest, and set the recursive exit condition in the recursive scheduling, that is, if the second standard deviation of the resource utilization rate of the same type of service virtual machine in each server is greater than or equal to the next balanced scheduling The second standard deviation of the resource utilization rate of the same type of service virtual machine in each server after the current balanced scheduling is performed, or as of the next balanced scheduling, the cumulative number of times the balanced scheduling is performed exceeds the first set threshold, then the second An executing unit 17 stops executing the acquisition of the second standard deviation of the resource utilization rate of the same type of service virtual machines in each server and the next balance scheduling. This helps to stop the above acquisition steps and balance scheduling in time, obtain the minimum standard deviation, and execute the balance scheduling corresponding to the minimum standard deviation.

Please refer to FIG. 6 , which is a schematic structural diagram of another distributed resource scheduling device provided by an embodiment of the present invention. The device 2000 includes:

The fifth acquiring unit 21 is configured to acquire the resource utilization ratio of various types of business virtual machines in each server in the distributed cluster system according to the ratio of various types of business virtual machines in each server to the resources of each server and the resource utilization rate of each type of business virtual machines in each server. The resource usage of each server.

Each server in the distributed cluster system has one or more types of business virtual machines, for example, there are desktop cloud virtual machines and compilation cloud virtual machines, and DRS allocates certain computing, storage, and input/output to each server in the cluster resources, and each server allocates the resources it obtains to various types of business virtual machines inside it according to a certain ratio. Ensuring that the total amount and ratio of resources of different types of business virtual machines in each server remain stable can ensure the service quality of various types of business virtual machines in each server.

For example, there are three servers in the cluster. Server 1 has 2 desktop cloud virtual machines and 2 compiled cloud virtual machines. Server 2 has 3 desktop cloud virtual machines and 2 compiled cloud virtual machines. Server 3 has desktop cloud. 1 virtual machine, 2 compiled cloud virtual machines, DRS allocates certain CPU, memory and I/O resources to servers 1, 2, and 3 in the cluster, and each server allocates certain CPU to each internal virtual machine , memory and I/O resources, taking CPU resource allocation as an example, DRS allocates CPU resources of quantity A to server 1, allocates CPU resources of quantity B to server 2, and allocates CPU resources of quantity C to server 3, and server 1 Allocate the CPU resources of A1, A2, A3 and A4 to the 2 desktop cloud virtual machines and 2 compiled cloud virtual machines respectively, and the resource utilization rates of the 2 desktop cloud virtual machines in server 1 are 30% and 40% respectively. %, the resource utilization rates of the two compiled cloud virtual machines in server 1 are 40% and 50% respectively, then the CPU resource utilization rate of server 1 is obtained (30%*A1+40%*A2+40%*A3+ 50%A4)/(A1+A2+A3+A4), and then calculate the memory and I/O resource usage of various virtual machines in server 1 in the same way, and then calculate the CPU, memory and I/O resource usage of server 1 The resource usage of I/O is weighted and calculated to obtain the resource usage of server 1, and so on for servers 2 and 3.

The sixth acquisition unit 22 is configured to acquire the third standard deviation of the resource utilization rate of the distributed cluster system through a predetermined acquisition method according to the resource utilization rate of each server, and the acquisition step is: according to the resource utilization rate of each server Obtain the average resource utilization rate of the distributed cluster system according to the resource utilization rate of the server, and obtain the resources of the distributed cluster system according to the resource utilization rate of each server and the average resource utilization rate of the distributed cluster system. The third standard deviation of utilization.

The difference from the foregoing embodiments is that since it is aimed at various types of business virtual machines in each server, the resource utilization rate of each server and the average resource utilization rate of the entire distributed cluster system are obtained in the same manner as the foregoing embodiments, and then , the sixth obtaining unit 22 obtains the standard deviation of the resource utilization rate of the distributed cluster system according to the formula (1).

The second scheduling unit 23 is configured to schedule any The service virtual machines are balanced and dispatched to other servers in the distributed cluster system.

Compare the obtained standard deviation of the resource utilization rate of the distributed cluster system with the scheduling threshold of the resource utilization rate of the distributed cluster system, if the standard deviation of the resource utilization rate of the distributed cluster system is greater than or equal to the distributed cluster system If the scheduling threshold of the resource usage of the system is exceeded, the second scheduling unit 23 triggers balanced scheduling. The object of balanced scheduling can be any service virtual machine in each server.

The seventh obtaining unit 24 is configured to obtain the resource utilization rate of each server in the distributed cluster system after balanced scheduling.

The eighth acquisition unit 25 is configured to acquire the fourth value of the resource utilization rate of the distributed cluster system after balanced scheduling through the predetermined acquisition method according to the resource utilization rate of each server in the distributed cluster system after balanced scheduling. standard deviation.

The second comparison unit 26 is configured to compare the fourth standard deviation obtained after at least one balance scheduling;

The second executing unit 27 is configured to execute the balanced scheduling corresponding to the smallest fourth standard deviation.

The functions of the seventh acquisition unit 24, the eighth acquisition unit 25, the second comparison unit 26, and the second execution unit 27 are respectively the same as those of the third acquisition unit 14, the fourth acquisition unit 15, the first comparison unit 16, the second execution unit 27 of the foregoing embodiment. The function of an execution unit 17 is similar, and will not be repeated here. The difference is that it is aimed at various types of business virtual machines in each server, and obtains the standard deviation of the resource utilization rate of the distributed cluster system.

It can be seen that according to the distributed resource scheduling device provided by the embodiment of the present invention, by performing distributed resource scheduling according to the resource ratio of various business virtual machines in each server to the server where it is located, the service of each business virtual machine in each server can be guaranteed. quality.

As a feasible implementation manner, the second execution unit 27 is also configured to if the fourth standard deviation of the resource utilization rate of the distributed cluster system after the next balanced scheduling is greater than or equal to the distributed cluster system after the current balanced scheduling. The fourth standard deviation of the resource utilization rate of the cluster system, or as of the next balanced scheduling, the cumulative number of times the balanced scheduling is performed exceeds the first set threshold, or the service virtual machine is balanced from any first server to After any second server, the change of the total amount of all kinds of business virtual machines in the any first server or the any second server is greater than the second set threshold, and/or the any first server Or if the change in the resources of any of the first servers occupied by various types of business virtual machines in any of the second servers is greater than the third set threshold, stop executing the resource utilization rate of the distributed cluster system after the balanced scheduling acquisition and the next balance scheduling.

Using recursive scheduling can make the scheduling efficiency the highest, and set the recursive exit condition in the recursive scheduling, that is, if the fourth standard deviation of the resource utilization rate of the distributed cluster system after the next balanced scheduling is greater than or equal to the current balanced scheduling The fourth standard deviation of the resource utilization rate of the distributed cluster system, or as of the next balanced scheduling, the cumulative number of times the balanced scheduling is performed exceeds the first set threshold, or the service virtual machine is switched from any first After the server is balanced and dispatched to any second server, the change of the total amount of various business virtual machines in the any first server or the any second server is greater than the second set threshold, and/or the any If the change in resources occupied by various types of service virtual machines in the first server or any second server is greater than the third set threshold, the second execution unit 27 stops executing the balanced scheduling Acquisition of the resource utilization rate of the post-distributed cluster system and the next balanced scheduling. This helps to stop the above acquisition steps and balance scheduling in time, obtain the minimum standard deviation, and execute the balance scheduling corresponding to the minimum standard deviation.

Please refer to FIG. 7, which is a schematic structural diagram of a distributed resource scheduling device provided by an embodiment of the present invention. The device 3000 of this embodiment includes a processor 31, a memory 32, an input device 33, an output device 34, and a bus system 35, wherein :

The processor 31 controls operations of the device 3000, and the processor 31 may also be called a central processing unit (Central Processing Unit, CPU). The processor 31 may be an integrated circuit chip with signal processing capabilities. The processor 31 may also be a general-purpose processor, a digital signal processor (Digital Signal Processing, DSP), an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field-programmable gate array (Field-Programmable Gate Array, FPGA) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. A general-purpose processor may be a microprocessor, or the processor may be any conventional processor, or the like.

The memory 32 may include read-only memory and random-access memory, and provides instructions and data to the processor 31 . A portion of memory 32 may also include non-volatile random access memory (NVRAM).

The various components of the device 3000 are coupled together through a bus system 35, which may be an Industry Standard Architecture (Industry Standard Architecture, ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (Extended Industry Standard Architecture, EISA) bus, etc. The bus can be one or more physical lines, and when there are multiple physical lines, it can be divided into an address bus, a data bus, a control bus, and the like. In some other embodiments of the present invention, the processor 31, the memory 32, the input device 33, and the output device 34 may also be directly connected through communication lines.

The input device 33 can be embodied as a mouse, keyboard, microphone, etc., and the output device 34 can be embodied as a display, an audio device, and a video device. Of course, the input device 33 and the output device 34 can also realize their functions through an input and output device, such as a touch screen.

Wherein, the processor 31 reads the computer program in the memory 32 to perform the following steps:

Obtain the resource utilization rate of the same type of business virtual machines in each server in the distributed cluster system;

According to the resource utilization rate of the same type of service virtual machine in each server, the first standard deviation of the resource utilization rate of the same type of service virtual machine in each server is obtained through a predetermined acquisition step, and the acquisition step is: Obtain the average resource usage rate of the same type of service virtual machines in each server according to the resource usage rate of the same type of service virtual machines in each server, and obtain the resource usage rate of the same type of service virtual machines in each server rate and the average resource usage rate, and obtain the first standard deviation of the resource usage rates of the same type of service virtual machines in each server;

When the first standard deviation of the resource usage rates of the same type of service virtual machines in each server is greater than or equal to the scheduling threshold of the resource usage rate of the same type of service virtual machines in each server, the distributed cluster The same type of business virtual machine of any server in the system is balancedly dispatched to other servers in the distributed cluster system;

Obtain the resource utilization rate of the same type of business virtual machines in each server in the distributed cluster system after balanced scheduling;

According to the resource utilization rate of the same type of service virtual machine in each server in the distributed cluster system after the balanced scheduling, obtain the resource utilization rate of the same type of service virtual machine in each server through the predetermined acquisition step the second standard deviation of

Comparing the second standard deviation obtained after the balanced scheduling at least once, and executing the balanced scheduling corresponding to the smallest second standard deviation.

As a feasible implementation manner, before the processor 31 performs the step of comparing the second standard deviation acquired after the balanced scheduling at least once, and performing the step of the balanced scheduling corresponding to the smallest second standard deviation , also perform the following steps:

If the second standard deviation of the resource utilization rate of the same type of service virtual machine in each server after the next balanced scheduling is greater than or equal to the resource utilization rate of the same type of service virtual machine in each server after the current balanced scheduling The second standard deviation of the second standard deviation, or as of the next balance scheduling, if the cumulative number of balanced scheduling exceeds the first set threshold, the acquisition step and the next balanced scheduling will be stopped.

The computer program contained in the processor 31 provided by the embodiment of the present invention can also be implemented as a first acquisition unit, a second acquisition unit, a first scheduling unit, a third acquisition unit, a fourth acquisition unit, a first comparison unit and a first For the execution unit, the functions implemented by these units can be referred to the foregoing embodiments, and will not be repeated here.

It can be seen that, according to a distributed resource scheduling device provided by an embodiment of the present invention, by performing distributed resource scheduling according to the types of service virtual machines in each server in a distributed cluster system, the service quality of each service virtual machine in each server can be guaranteed. .

Please refer to FIG. 8, which is a schematic structural diagram of another distributed resource scheduling device provided by an embodiment of the present invention. The device 4000 of this embodiment includes a processor 41, a memory 42, an input device 43, an output device 44, and a bus system 45. in:

The processor 41 controls operations of the device 4000, and the processor 41 may also be referred to as a central processing unit (Central Processing Unit, CPU). The processor 41 may be an integrated circuit chip with signal processing capability. The processor 41 can also be a general-purpose processor, a digital signal processor (Digital Signal Processing, DSP), an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field-programmable gate array (Field-Programmable Gate Array, FPGA) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. A general-purpose processor may be a microprocessor, or the processor may be any conventional processor, or the like.

The memory 42 may include read-only memory and random-access memory, and provides instructions and data to the processor 41 . A portion of memory 42 may also include non-volatile random access memory (NVRAM).

The various components of the device 4000 are coupled together through a bus system 45, which may be an Industry Standard Architecture (Industry Standard Architecture, ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (Extended Industry Standard Architecture, EISA) bus, etc. The bus can be one or more physical lines, and when there are multiple physical lines, it can be divided into an address bus, a data bus, a control bus, and the like. In some other embodiments of the present invention, the processor 41, the memory 42, the input device 43, and the output device 44 may also be directly connected through communication lines.

The input device 43 can be embodied as a mouse, a keyboard, a microphone, etc., and the output device 44 can be embodied as a display, an audio device, and a video device. Of course, the input device 43 and the output device 44 can also realize their functions through an input and output device, such as a touch screen.

Wherein, the processor 41 reads the computer program in the memory 42 to perform the following steps:

According to the proportion of various service virtual machines in each server in the distributed cluster system to the resources of each server and the resource utilization rate of various service virtual machines in each server, the resource utilization rate of each server is obtained;

According to the resource utilization rate of each server, the first standard deviation of the resource utilization rate of the distributed cluster system is acquired through a predetermined acquisition step, and the acquisition step is: acquiring the resource utilization rate of each server according to the the average resource utilization rate of the distributed cluster system, and obtain the first standard deviation of the resource utilization rate of the distributed cluster system according to the resource utilization rate of each server and the average resource utilization rate of the distributed cluster system;

When the first standard deviation of the resource usage rate of the distributed cluster system is greater than or equal to the scheduling threshold of the resource usage rate of the distributed cluster system, any service virtual machine in each server is balancedly scheduled to the In other servers of the distributed cluster system;

Obtain the resource utilization rate of each server in the distributed cluster system after balanced scheduling;

According to the resource usage rate of each server in the distributed cluster system after the balanced scheduling, the second standard deviation of the resource usage rate of the distributed cluster system after the balanced scheduling is obtained through the predetermined acquisition step;

Comparing the second standard deviation obtained after the balanced scheduling at least once, and executing the balanced scheduling corresponding to the smallest second standard deviation.

As a feasible implementation manner, before the processor 41 compares the second standard deviation acquired after the balanced scheduling at least once, and executes the step of the balanced scheduling corresponding to the smallest second standard deviation, Also perform the following steps:

If the second standard deviation of the resource usage rate of the distributed cluster system after the next balanced scheduling is greater than or equal to the second standard deviation of the resource usage rate of the distributed cluster system after the current balanced scheduling, or until the next time Balanced scheduling, the cumulative number of times the balanced scheduling is performed exceeds the first set threshold, or after the service virtual machine is balancedly scheduled from any first server to any second server, any one of the first servers or all The change of the total amount of various types of business virtual machines in any second server is greater than the second set threshold, and/or the proportion of various types of business virtual machines in any one of the first servers or the second server is If the resource change of any one of the first servers is greater than the third set threshold, the execution of the acquisition step and the next balance scheduling will be stopped.

The computer program included in the processor 41 provided by the embodiment of the present invention can also be implemented as a fifth acquisition unit, a sixth acquisition unit, a second scheduling unit, a seventh acquisition unit, an eighth acquisition unit, a second comparison unit, and a second For the execution unit, the functions implemented by these units can be referred to the foregoing embodiments, and will not be repeated here.

It can be seen that according to the distributed resource scheduling device provided by the embodiment of the present invention, by performing distributed resource scheduling according to the resource ratio of various business virtual machines in each server to the server where it is located, the service of each business virtual machine in each server can be guaranteed. quality.

It should be noted that for the foregoing method embodiments, for the sake of simple description, they are expressed as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described action sequence. Because according to the present invention, certain steps may be performed in other order or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification belong to preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

In the foregoing embodiments, the descriptions of each embodiment have their own emphases, and for parts not described in detail in a certain embodiment, reference may be made to relevant descriptions of other embodiments.

Through the above description of the implementation manners, those skilled in the art can clearly understand that the present invention can be implemented by hardware, firmware, or a combination thereof. When implemented in software, the functions described above may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media may be any available media that can be accessed by a computer. Take this as an example but not limited to: computer-readable media may include Random Access Memory (Random Access Memory, RAM), Read-Only Memory (Read-Only Memory, ROM), Electrically Erasable Programmable Read-Only Memory (Electrically Erasable Programmable Read -Only Memory, EEPROM), CD-ROM (Compact Disc Read-Only Memory, CD-ROM) or other optical disk storage, magnetic disk storage medium or other magnetic storage devices, or can be used to carry or store information in the form of instructions or data structures desired program code and any other medium that can be accessed by a computer. also. Any connection can suitably be a computer-readable medium. For example, if the Software is transmitted from a website, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, Digital Subscriber Line (DSL), or wireless technologies such as infrared, radio, and microwave, Coaxial cable, fiber optic cable, twisted pair, DSL or wireless technologies such as infrared, wireless and microwave are then included in the fixation of the associated medium. As used herein, disk and disc include compact disc (CD), laser disc, compact disc, digital versatile disc (DVD), floppy disc, and Blu-ray disc, where a disc usually reproduces data magnetically, and a disc Lasers are used to optically reproduce the data. Combinations of the above should also be included within the scope of computer-readable media.

In a word, the above descriptions are only preferred embodiments of the technical solutions of the present invention, and are not intended to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims (8)

1. a kind of distributed resource scheduling method, it is characterised in that including：

Obtain the resource utilization of the same type business virtual machine in distributed cluster system in each server；

According to the resource utilization of the same type business virtual machine in each server, obtained by predetermined obtaining step First standard deviation of the resource utilization of the same type business virtual machine in each server, the obtaining step is：Root The same type in each server is obtained according to the resource utilization of the same type business virtual machine in each server The average resource utilization rate of business virtual machine, and used according to the resource of the same type business virtual machine in each server The resource utilization of same type business virtual machine in rate and the average resource utilization rate, acquisition each server First standard deviation；First standard deviation is used for the resource utilization for reacting same type business virtual machine in each server Dispersion degree；

When the first standard deviation of the resource utilization of the same type business virtual machine in each server is more than or equal to institute During the scheduling thresholds for the resource utilization for stating same type business virtual machine in each server, using Recursive schedule mode by institute The same type business virtual machine balance for stating any server in distributed cluster system is dispatched to distributed cluster system Other servers in；

Obtaining the resource of the same type business virtual machine after balance scheduling in distributed cluster system in each server makes With rate；

According to the money of the same type business virtual machine in distributed cluster system after the balance scheduling in each server Source utilization rate, the resource for obtaining the same type business virtual machine in each server by the predetermined obtaining step makes With the second standard deviation of rate；

Compare second standard deviation obtained after the balance scheduling at least one times, perform minimum second standard deviation pair The balance scheduling answered.

2. the method as described in claim 1, it is characterised in that described to compare the institute obtained after the balance scheduling at least one times The second standard deviation is stated, before the corresponding balance scheduling of second standard deviation for performing minimum, in addition to：

If carrying out the of the resource utilization of the same type business virtual machine after balance scheduling next time in each server Two standard deviations are more than or equal to the resource for carrying out same type business virtual machine after forward horizontal stand is dispatched in each server Second standard deviation of utilization rate, or end balance scheduling next time, the cumulative number for carrying out the balance scheduling is set more than first Determine threshold value, then stop performing the obtaining step and balance scheduling next time.

3. a kind of distributed resource scheduling method, it is characterised in that including：

According in distributed cluster system in each server all kinds of business virtual machines account for the resource of each server ratio and The resource utilization of all kinds of business virtual machines in each server, obtains the resource utilization of each server；

According to the resource utilization of each server, the money of the distributed cluster system is obtained by predetermined obtaining step 3rd standard deviation of source utilization rate, the obtaining step is：The distribution is obtained according to the resource utilization of each server The average resource utilization rate of formula group system, and the resource utilization according to each server and the distributed type assemblies system The average resource utilization rate of system, obtains the 3rd standard deviation of the resource utilization of the distributed cluster system；3rd mark Quasi- difference is used for the dispersion degree for reacting the resource utilization of the distributed cluster system；

When the 3rd standard deviation of the resource utilization of the distributed cluster system is more than or equal to the distributed cluster system Resource utilization scheduling thresholds when, any business virtual machine balance in each server is adjusted using Recursive schedule mode Degree is into other servers of the distributed cluster system；

Obtain the resource utilization of each server in distributed cluster system after balance scheduling；

According to the resource utilization of each server in distributed cluster system after the balance scheduling, pass through the predetermined acquisition Step obtains the 4th standard deviation of the resource utilization of distributed cluster system after the balance scheduling；

Compare the 4th standard deviation obtained after the balance scheduling at least one times, perform minimum the 4th standard deviation pair The balance scheduling answered.

4. method as claimed in claim 3, it is characterised in that described to compare the institute obtained after the balance scheduling at least one times The 4th standard deviation is stated, before the corresponding balance scheduling of the 4th standard deviation for performing minimum, in addition to：

If the 4th standard deviation for carrying out the resource utilization of the distributed cluster system after balance scheduling next time is more than or waited Put down next time in the 4th standard deviation for carrying out the resource utilization of distributed cluster system after forward horizontal stand is dispatched, or cut-off Weighing apparatus scheduling, carries out the cumulative number of the balance scheduling more than the first given threshold, or by the business virtual machine from any the One server balance is dispatched to after any second server, each in any first server or any second server The change of the total amount of class business virtual machine is more than the second given threshold, and/or any first server or described any the In two servers all kinds of business virtual machines account for the resource of any first server change be more than the 3rd given threshold, then Stop performing the obtaining step and balance scheduling next time.

5. a kind of distributed resource scheduling device, it is characterised in that including：

First acquisition unit, the resource for obtaining the same type business virtual machine in distributed cluster system in each server Utilization rate；

Second acquisition unit, for the resource utilization according to the same type business virtual machine in each server, passes through Predetermined acquisition modes obtain the first standard deviation of the resource utilization of the same type business virtual machine in each server, The acquisition modes are：Each clothes are obtained according to the resource utilization of the same type business virtual machine in each server The average resource utilization rate for the same type business virtual machine being engaged in device, and according to the same type business in each server The resource utilization of virtual machine and the average resource utilization rate, the same type business obtained in each server are virtual First standard deviation of the resource utilization of machine；It is empty that first standard deviation is used to react same type business in each server The dispersion degree of the resource utilization of plan machine；

First scheduling unit, the first mark for the resource utilization when the same type business virtual machine in each server During the scheduling thresholds of quasi- resource utilization of the difference more than or equal to the same type business virtual machine in each server, use The same type business virtual machine of any server in the distributed cluster system is balanced and dispatched by Recursive schedule mode Into other servers of distributed cluster system；

3rd acquiring unit, for obtaining the same type industry after balance scheduling in distributed cluster system in each server The resource utilization of business virtual machine；

4th acquiring unit, for according to the same class in distributed cluster system after the balance scheduling in each server The resource utilization of type business virtual machine, the same type industry in each server is obtained by the predetermined acquisition modes Second standard deviation of the resource utilization of business virtual machine；

First comparing unit, for comparing second standard deviation obtained after the balance scheduling at least one times；

First execution unit, the corresponding balance scheduling of second standard deviation for performing minimum.

6. device as claimed in claim 5, it is characterised in that first execution unit is additionally operable to：

If carrying out the of the resource utilization of the same type business virtual machine after balance scheduling next time in each server Two standard deviations are more than or equal to the resource for carrying out same type business virtual machine after forward horizontal stand is dispatched in each server Second standard deviation of utilization rate, or end balance scheduling next time, the cumulative number for carrying out the balance scheduling is set more than first Determine threshold value, then stop performing the second standard deviation of the resource utilization of the same type business virtual machine in each server Obtain and balance scheduling next time.

7. a kind of distributed resource scheduling device, it is characterised in that including：

5th acquiring unit, for according to all kinds of business virtual machines account for each service in each server in distributed cluster system The resource utilization of all kinds of business virtual machines in the ratio of the resource of device and each server, obtains the money of each server Source utilization rate；

6th acquiring unit, for the resource utilization according to each server, obtains described by predetermined acquisition modes 3rd standard deviation of the resource utilization of distributed cluster system, the acquisition modes are：According to the resource of each server Utilization rate obtains the average resource utilization rate of the distributed cluster system, and according to the resource utilization of each server with And the average resource utilization rate of the distributed cluster system, obtain the 3rd of the resource utilization of the distributed cluster system Standard deviation；3rd standard deviation is used for the dispersion degree for reacting the resource utilization of the distributed cluster system；

Second scheduling unit, the 3rd standard deviation for the resource utilization when the distributed cluster system is more than or equal to institute , will be any in each server using Recursive schedule mode during the scheduling thresholds for the resource utilization for stating distributed cluster system Business virtual machine balance is dispatched in other servers of the distributed cluster system；

7th acquiring unit, the resource utilization for obtaining each server in distributed cluster system after balance scheduling；

8th acquiring unit, for the resource utilization according to each server in distributed cluster system after the balance scheduling, The 4th standard of the resource utilization of distributed cluster system after the balance scheduling is obtained by the predetermined acquisition modes Difference；

Second comparing unit, for comparing the 4th standard deviation obtained after the balance scheduling at least one times；

Second execution unit, the corresponding balance scheduling of the 4th standard deviation for performing minimum.

8. device as claimed in claim 7, it is characterised in that second execution unit is additionally operable to：

If the 4th standard deviation for carrying out the resource utilization of the distributed cluster system after balance scheduling next time is more than or waited Put down next time in the 4th standard deviation for carrying out the resource utilization of distributed cluster system after forward horizontal stand is dispatched, or cut-off Weighing apparatus scheduling, carries out the cumulative number of the balance scheduling more than the first given threshold, or by the business virtual machine from any the One server balance is dispatched to after any second server, each in any first server or any second server The change of the total amount of class business virtual machine is more than the second given threshold, and/or any first server or described any the In two servers all kinds of business virtual machines account for the resource of any first server change be more than the 3rd given threshold, then Stop perform it is described balance scheduling after distributed cluster system resource utilization acquisition and next time balance scheduling.